Out of service escalator at Westlake Station

David Cole has a largely correct takedown of Sound Transit’s traditional pound-foolish approach to escalators:

Wherever long escalators are required to travel between the train platform and street level, redundant escalators should be provided. This could have been accomplished with a single bank of at least four escalators, or two banks of at least three escalators, etc. With a bank of four escalators, one escalator being out of service would be a minor inconvenience at worst. Even with half the escalators out of service, access to and from the station could be maintained …

Beyond the number of escalators at each station, there is also the issue of the escalators themselves. Broken escalators have been a near-constant bane to riders using the Capitol Hill and UW stations since their opening in 2016, to the point that Sound Transit is already planning to replace all 13 escalators at UW less than four years after that station’s completion.

But Sound Transit apparently felt differently back when planning the Capitol Hill and UW stations. According to a source familiar with the design process who declined to be named for this article, Sound Transit insisted on specifying medium-duty “better” escalators at these stations as a cost-saving measure, and then cobbled together a myriad series of customizations to bring them up to heavy-duty standards. As we now know, the reliability of these Frankenstein escalators hasn’t exactly been stellar, and Sound Transit will soon spend a fortune to replace them with more robust, off-the-shelf models. Some old adage comes to mind about how it’s better to do something right the first time than to do it over again.

The nicest thing to say here is that the highest levels now appreciate that this is a serious problem, even if the rest of the agency isn’t quite ready to take full ownership of it.

The more useful thing to add is that center platforms provide intrinsic redundancy. For the same number of elevators and escalators, any rider has twice as many options for getting to the platform.

43 Replies to “Center platforms are the inexpensive escalator design”

  1. Maybe they learned their lesson with elevators?

    I don’t read about the Link associated elevators having issues. But the ones along SR520 are down almost every week now, and the only option is to take stairs. As long as you can do that.

    Just dumb.

    1. Fortunately, most of the people using the bus stops along 520 are simply transferring between buses serving the same stop. Even if access to the surrounding neighborhood were cut off completely, they wouldn’t notice.

    2. Jack Nolan, the elevators are no better. Seatac’s elevator from the pedestrian walkway to street level is about to be shut down for 2 months and be completely overhauled. The Kone elevators have a worse track record than the escalators do.

      1. I still have never understood why they never bothered to put in a second elevator along with a escalator bank there. They have this very large often empty pedestrian plaza and parking lot loop that could be better served with adding said features.

  2. If I’m not mistaken, both the UW and Capitol Hill Stations are center platform. Both have had problems. The biggest issue is lack of stairs.

    1. UWS has allowed access to the stairs for at least a month. I use them all the time…it’s an interesting change of scenery.

    2. I use the stairs at UW/Husky for a healthy workout. All seven landings worth! And they are a lot less crowded than the escalators–LOL. A recent sign at the Capitol Hill station said one stairwell would soon be open for use from the platform to the mezzanine.

  3. Although not ideal, ST could install track crossings at the ends of platforms. ST has pedestrian-only double track crossings planned for East Main and SE Redmond. Maybe the track crossings are fenced and opened only when there is an out-of-service elevator or escalator. Sure there is a safety risk but it could be a quick and easy interim solution.

    1. Sight lines in tunnels won’t allow that. In the at-grade settings, the operator can see the whole area approaching a station. This is not the case in a tunnel. Plus its just really bad safety practice to encourage people to enter an active railroad.

      1. There are a few above-ground places where this could work because a train operator would have plenty of time to see and react. An example: a crossing just south of the Mt Baker platforms that could then continue as a pedestrian overcrossing for Rainier and MLK (replacing the current one.

        For that matter, both East Main and SE Redmond appear to have sight distance issues. Easy Main’s crossing is just outside of a tunnel!

  4. ST designs stations without putting in placeholders for future escalators or elevators. It’s penny-wise and pound-foolish to do this. Unfortunately, all of the new stations now being designed or constructed appear to disregard future allowances for these —even when the items are removed at the last minute for cost-cutting.

  5. I read the UW/Husky station replacement plan again. It says they are replacing 11 of 13 escalators. Two will be replaced with stairs. That leads me to believe no down escalators will reach the platform. Is that a correct assumption?

    1. Yes, ST announced the bottommost down escalators would be replaced with stairs. It’s another quality degradation. You don’t see department stores having no escalator to the second floor because they have to attract customers, but with transit infrastructure it’s a captive market because you can’t go to another subway with better circulation.

  6. I find it shameful that ST will ask local cities to come up with extra hundreds of millions for rail tunnels — but not as local cities to come up with a few extra million for more escalators and elevators. If there has to be cost-cutting, shouldn’t ST at least ask the cities this before unilaterally removing them?

    This is exactly what happened recently with the Lynnwood Link cost-cutting. There was no public discussion about escalator removal.

    1. “This is exactly what happened recently with the Lynnwood Link cost-cutting. There was no public discussion about escalator removal.”

      Bingo. I’ve brought this issue to the Lynnwood Link project team’s attention at least three times, the most recent being at the open house ST held a couple of months ago at the Lynnwood Convention Center. I’m still waiting for a reply.

  7. “Broken escalators have been a near-constant bane to riders using the Capitol Hill and UW stations”

    UW Station only. Capitol Hill has had a few escalator breakdowns but not even a quarter of UW’s. I use both stations several times a week, and the last time I saw a broken escalator at Capitol Hill was several months ago. (It was the south platform up, which I think has broken a couple times.)

    In contrast, UW had two years where it couldn’t go three days without one escalator somewhere being out of service, or even multiple ones. I’ve had times where the surface down was closed on one side so I went to the other side, then the long down was closed on that side and I had to walk around on the mezzanine. If I had known that I would have taken the elevator but it was too late by then.

    And this past week the north surface down was closed for two days, and then both of the north surface escalators were closed the next day, and one of the elevators was closed for at least two days. I was hoping maybe they were finally replacing the north surface escalators, but maybe not yet.

    1. At least now, if you get to the mezzanine level and find broken down escalators, the stairs are a conveniently placed option. It’s not a horrible alternative, and I enjoy the flexibility of choice.

      It’s weird how the breakdown eras have been cyclical…I mean, at opening it was ok, then for a period after it was an everyday breakdown/being broken, finally a year or so ago they seemed to get things into the more-working-than-not lane…which has again devolved into mostly something being broken. What were they doing differently that worked for that previous “mostly working” period?

      I’m so curious what the primary cause of breakdowns is? People dropping things into the mechanism? Too much weight? Pooping while riding (that was actually mentioned in an earlier ST report)? Or is it just boring normal wear-and-tear?

    2. Parts that were rated to last thirty years wore out in one year. The warranty period began when they started testing the elevators before opening, so they were just out-of-warranty when the problems started happening. ST lowballed the contact spec by requiring only moderate-use escalators rather than heavy-duty ones and not requiring a longer-term warranty. That was supposed to save money. Then on top of that are the universal problems of people tripping on a step or getting a shoelace or scarf caught in it. When passengers started complaining persistently ST merely said, “All transit agencies have escalator problems.” Not. Like. This. The Mt Baker, TIB, SeaTac, and DSTT escalators never had this many problems. A DSTT escalator sometimes closed for months while the county waited for a federal grant to fund repairs, but it didn’t close repeatedly three times a week.

  8. The other day at the Seatac Link station both downward/egress escalators were out of service. There were two up escalators and the elevators.

    So where exactly is this intrinsic redundancy inherent to center platforms?

    1. Center platforms are better all around. It doubles the number of exits and elevators available. It avoids going up to the surface to transfer to an opposite-direction train, or if you missed your stop or turn around en route or walked down to the wrong platform. I’ve had multiple times where I’ve automatically walked to the platform I usually use only to remember I’m going the other direction today, or I’ll start on a trip but it’s taking so long that I decide to go to another destination instead and turn around or go back home. Center platforms make that easy. Side platforms add unnecessary barriers, and no advantage to passengers. Center platforms may cost more because the track has to separate for the platform (or that was the reason they gave when they built the DSTT), but if we’re going to build it we should build it right the first time.

  9. I rarely used station elevators before UW Station opened; I wanted to save electricity and leave room for the disabled and bicyclists who need them. But UW Station broke my resolve. Now when I come across the bridge I look to see if any of the escalator entrances are closed and if so I head to the elevator. And I’ve been burned enough times with going down to the mezzanine and finding the escalator closed, or finding it closed multiple days, that even the day after an escalator was closed I sometimes use the elevator because I just find it really annoying to have a closed escalator sprung on you partway down. Likewise I won’t use the surface stairs because the escalators should be functioning. So ST has lost its passenger good-citizenship of staying away from the elevators because of ST’s own neglect of the escalators.

    1. I tried the elevators a couple times at UW Station, but they have got to be some of the slowest elevators imaginable. Despite having two elevators and only three floors, I spent over five minutes waiting for one to show up, and then it took over two minutes to get down to the bottom. Part of that was it was packed and we had to shuffle around to let people out at the ground level (I started on the second level) and part of it is just that they move sloooowly and the doors open/close slooooowly.

      It’ll be escalators or stairs from now on, unless I have my bike.

  10. One other discussion that is needed is that as Link expands, most existing stations will have more riders going through them. This is especially true for the existing four DSTT Stations — which all have side platforms and opened in 1990 (29 years ago and before ADA). Revisiting circulation for these stations is badly needed. Of course , two are going to have new platforms as required by ST3 — but Station overhauls for all four should be funded as an impact for all of these new riders to be brought in on four-car trains every four minutes on each side of each station. I even shudder to think what happens inside a station during minor service disruptions that result in a train going out of service. Other systems revisit existing stations when extensions open; BART added faregates in Downtown San Francisco through the SFO extension, for example. ST has yet to publicly state how many boarding will be at each station once the system opens.

    1. They should really add center platforms to the DSTT stations. Now that the buses are out there is no need for the center lane, and a center platform could expand capacity and improve station flow without a major rework of the existing station. They could probably even be added without closing the stations at all.

      1. But then how would you get off the center platform? Stairs, escalators, or elevators would need to be added. There may not be room for that.

      2. The escalators would need to be moved and new elevators installed. That would cost money but there’s plenty of space. The new platform could be the width of the two existing platforms, and there’s the extra median lane available. The problem is the cost and the disruption of shutting down Link or splitting it during construction. (There would have to be a path for northern trains to the SODO base.) ST is already planning to split the line for a few weeks in preparation for ST2, single-tracking both north and south of a station and requiring all passengers to make a cross-platform transfer on a temporary center platform to continue north or south. I don’t remember which station that is, Pioneer Square or University Street probably? So ST already has some ideas on how to do it; it just needs the will. The cost can be put down to fixing a broken design, the same way ST is replacing the substandard escalators.

      3. Concur 100%.

        ST would need to add ingress and egress to the center platform, which could be expensive, but it could be done, and it probably should be done at least at those stations where ST expects a lot of interline transfers to occur.

        And don’t forget the inherent advantages of being able to use the Spanish Solution.

        With buses out of the DSTT ST doesn’t need to plan for breakdowns now. Do something with that space.

  11. One other eventual “A! HA!” screw-up will come when ST will reveal and the public will discover that both East Main and Bellevue Downtown stations have side platforms. Riders from Eastgate/Issaquah will have to go all the way to Wilburton to get to a center platform once that Link line opens. For those going to Seattle, that will be unacceptable.

    Of course, the addition of branch tracking for Issaquah Link itself will be so contentious that I expect the line will be branched from the center-platform South Bellevue Station and across Mercer Slough there.

    1. So if you’re going from Issaquah to SeaTac you’ll have to make an out-of-platform transfer at Bellevue TC and again at Intl Dist.

      1. I’d just cross the tracks at East Main as designed rather than ride into Bellevue, change levels twice, then ride back.

        International District is a different beast altogether and as we all know it is extremely critical to get that transfer setup exactly right. They need to perform the thought exercise you just described – “if I’m on East Link and transferring to the airport, what is the exact route through the station I need to take to transfer?” – and do so for any possible permutation of transfers for any possible station design. Based on past history and the fact that there is still a possibility of an exceedingly deep component to the station for the ST3 lines, I am holding out very little hope – and unfortunately the depth of that station will probably cost any neighboring businesses far more in opportunity costs (“hey, I think I’ll just pop up and eat/shop here”) than they would ever lose during construction; local residents will lose as well as that deep station will forever be part of their transit experience instead of something that makes sense.

      2. Issaquah to SeaTac…even after ST3 Link is fully built-out, for most, this looks like a 4-seat ride: 1) car->Issaquah Transit Center, 2) train to East Main St. Station, 3) Train to International District Station, 4) Train to airport.

        On top of the 4-seat ride, you’re looking at a quite circuitious route – you’re going north, then back south. Then, after going west, the train jogs east to the Rainier Valley, then back west again.

        Back-of-the-envelope math to figure out the travel time of all this. Link from ID Station to SeaTac is currently scheduled at 31 minutes. Add 2 more for the new stops at Graham St. and BAR, you’re up to 33. Bellevue Link from East Main St. Station to ID Station is projected to be another 25, which brings the total to 58. The Issaquah Link segment needs to cover 12 miles with 2 intermediate stops, so I’ll say 15 minutes. That’s 73 minutes so far. Add 10 more for the local drive from home to Issaquah Transit Center, the total is now 83. But, we still haven’t accounted for any of the wait times or transfer overhead. Add an additional 5 minutes at each train station (assuming the initial car segment has zero wait time), we’re up to 98 minutes for the grand total. By contrast, Google Maps estimates a drive time of 26 minutes (18.7 miles) and, even today’s best transit option (554->Link) is estimated by Google at 1 hour 22 minutes, a full 16 minutes faster than the future car->train->train->train option.

        Overall, 1 hour 38 minutes feels like it’s too long to get people out of their cars (or Uber/Lyft cars). Even among the transit space, I can’t help wondering if come 2041, there will be better options, involving buses. One potential option could be to stay on Issaquah Link one stop further to Bellevue Transit Center, then board the I-405 BRT bus, switching to Link at Tukwila Station. I think it would be faster, but it’s still a 4-seat ride, and a somewhat circuitious one at that, going north to Bellevue Transit Center, only to go back south again.

        For another option, the King County Metro Long Range Plan shows an express route from Issaquah to Renton via SR-900. Unfortunately, after arriving in Renton, you still have to transfer to I-405 BRT for one stop, only to transfer again at Tukwila Station, so it doesn’t save any transfers. But, it at least makes for a route that’s more direct and not subject to traffic delays associated with 405. Plus, if the Metro route ends up covering more of Issaquah than just the transit center, perhaps (if you’re lucky), that first leg can be eliminated. Of course, the tradeoff is that this route probably won’t be all that frequent – likely, half-hourly at best, and it might not run at all if it’s an evening or weekend.

        But, at the end of the day, for someone actually transferring between all three Link lines to do Issaquah->SeaTac, I feel like the up and down to move between platforms is the least of their problems.

      3. Agreed, asdf2 – I honestly don’t see a lot of Issaquah-to-Airport transit use. It just isn’t cost or time effective. Where you might see a reasonably large number of transfers (with the caveat that Issaquah will never provide a huge number of passengers) is Issaquah-to-Seattle at East Main – and that assumes that just taking the bus to Mercer Island and transferring there isn’t faster/easier.

      4. It may not be a lot of people, but for people who do want to take transit to the airport it makes a difference. And we’re not asking for an expensive custom route, we’re just asking for the biggest transfer points in the network to be designed well for transfers.

    2. East Main is at-grade with at-grade crossings between platforms, so not as onerous as having to change levels. Bellevue Downtown is worse because you do have to change levels, and there’s no passenger-oriented reason why you couldn’t directly access a center platform from either entrance. They were likely saving money because the tracks exit the tunnel there and they’d have to widen it a bit to allow the tracks to diverge. There’d also be some additional guideway costs as the guideway would have to be split until it merged east of the station.

      My guess is that they’re using the side platforms at East Main because there is no need for a mezzanine (the largest additional cost to a center-platform station in most cases) and because using a center platform would require everyone using the station to cross the tracks, not just the westbound passengers.

      Anybody transferring or reversing direction would do so at East Main as it’s straightforward there.

      1. Certainly East Main transfers will be level and that’s better than stairs. Still, the creation of side platforms means that transferring riders will have to walk up to about 300 feet rather than 20 feet across a center platform.

      2. True, although people frequently making transfers there will gravitate towards using the first/last cars (yet another reason for open gangway cars, grrrr) so that they can cross quickly to the other platform. While I agree that center platforms are generally better, in this case I can understand why they didn’t provide one.

        Fully agree with other posters that center platforms should be added at downtown stations even (especially) if they are basically meant for transfers only, Spanish solution or otherwise. Elevators are likely sufficient to meet ADA, with the addition of gated “emergency only” cross-track crossings at each end of the platform. Train announcements would be something like “Exit to left for transfers to lines A, B, C; exit to right to leave station.” This would also provide the immense advantage of a center platform having direct access down to a perpendicular center platform on the Ballard Link line below.

    3. They better end up branching at I-90 to Issaquah. Its absolutely ridiculous to build duplicative track all the way from I-90 to Downtown Bellevue to avoid a few hundred feet of crossing a swamp that is already covered with 10 lanes of highway ramps. ST needs to fight these No Mercer Slough NatureNazis, afterall the cost savings to ST are huge, not to mention the huge time savings for riders.

  12. As a consumer of rail transit I much prefer center platforms, but I wouldn’t say that the are cheaper than side platforms because of the potential for reduced escalator and ingress/egress. In fact, they can be significantly more expensive.

    1. Center platforms make the most sense where two lines meet and riders transfer in the opposite direction. Of the multiple transfer stations planned for Link, only Wilburton fits this basic operational strategy. The other five do not.

      If the Fifth shallow station option is chosen, ID may have this in one direction. As noted elsewhere, East Main is at least level. The final layout of Tacoma Dome and SoDo have not been finalized by early concepts seem to all require level changes. I’m not sure if a small group could change the design, but if enough advocates would lobby for it, changes could happen.

  13. I’ve been using a cane due to an accident. I now have a lot of pity for disabled riders that use Sound Transit. University St Station is like a maze to get in and out of with multiple escalators and elevators down. Capitol Hill station is a little better.
    UW station is ridiculous. Escalators have been down for weeks. Elevators are inconvenient, and also not always working. The design of these stations seems like a big middle finger to the disabled.

  14. For long vertical distances, elevators are much preferred. That UW Station escalator ride is nearly 3 minutes, but maybe 20 seconds on the elevator. Yes, install top-grade equipment, ST, with optimum maintenance, but definitely more elevators for deep tunnel stations.

  15. the platform level to mezzanine upper level do not have stairs at capital hill station and university of Washington station far as I know. I have never found the stairs to get from the platform level to mezzanine upper level. Their is stairs from the mezzanine upper level to the street. Also the mezzanine upper level does not connect from each side of the escalators. You have to make the decision on the platform level on which side of the station you will end up when you get to the street. The only station that the mezzanine level connects all stairs, escalators, and elevators from platform level is at West Lake station. All stations mezzanine level should connect to all stairs, escalators, and elevators that go to the platform level. All stations should have stairs that go to all levels of the station.

Comments are closed.